Year: 2018
Author: Howard, Sarah, Gonzalez, Carlos, Montenegro, Maximilano, Yang, Jie, Jun, Ma
Type of paper: Abstract refereed
Abstract:
Over the last two decades, we have seen an influx of digital technologies in education. There have also been significant changes in educational research. Specifically, the fields of Educational Data Mining (EDM) and Learning Analytics (LA), coming from computer science roots, have developed into active fields of research applying data mining techniques and traditional statistical approaches to examine educational practice, students learning paths and patterns in educational data, etc.. The overall aim of both fields has been to improve educational practice and inform learning. In the area of educational technology research, these approaches are particularly useful because data can be gathered from a range of online platforms and digital tools used in learning.
In this paper, we present a close study of how data mining approaches can build on and extend traditional statistical findings, to gain a more complete picture of students’ engagement with online resources. The aim is to encourage the use of data mining approaches in educational research, beyond the field of EDM. Traditional statistics will usually seek to deductively fit a parameterized model. Data mining approaches are more inductive, where a model is created from a dataset and then tested for accuracy, to reveal new relationships to explain phenomena and patterns.
To present this, we reanalyze data from a study examining factors of online resource use contributing to student achievement. The dataset comprises library loan data, electronic journal access and student academic information (Montenegro et al., 2016). We created ‘fuzzy sets’ and conducted an association rules analysis. Montenegro et al. had identified two groups of students based on accessing electronic library resources and that access changed over time. The new analysis identified that accessing electronic library resources only contributed to students’ high achievement. Lower performing students were more likely to be affected by the number of years in university, but not from their rate of accessing electronic library resources. These findings build on the original analysis to possibly explain changes in students use of library resource over time, as they spent more time in university. Implications of the combined findings will be discussed and future directions for this approach will be explored.
Montenegro, M., Clasing, P., Kelly, N., Gonzalez, C., Jara, M., Alarcón, R., … aurina, E. (2016). Library Resources and Students’ Learning Outcomes: Do All the Resources Have the Same Impact on Learning? The Journal of Academic Librarianship, 42(5), 551–556. https://doi.org/10.1016/J.ACALIB.2016.06.020
In this paper, we present a close study of how data mining approaches can build on and extend traditional statistical findings, to gain a more complete picture of students’ engagement with online resources. The aim is to encourage the use of data mining approaches in educational research, beyond the field of EDM. Traditional statistics will usually seek to deductively fit a parameterized model. Data mining approaches are more inductive, where a model is created from a dataset and then tested for accuracy, to reveal new relationships to explain phenomena and patterns.
To present this, we reanalyze data from a study examining factors of online resource use contributing to student achievement. The dataset comprises library loan data, electronic journal access and student academic information (Montenegro et al., 2016). We created ‘fuzzy sets’ and conducted an association rules analysis. Montenegro et al. had identified two groups of students based on accessing electronic library resources and that access changed over time. The new analysis identified that accessing electronic library resources only contributed to students’ high achievement. Lower performing students were more likely to be affected by the number of years in university, but not from their rate of accessing electronic library resources. These findings build on the original analysis to possibly explain changes in students use of library resource over time, as they spent more time in university. Implications of the combined findings will be discussed and future directions for this approach will be explored.
Montenegro, M., Clasing, P., Kelly, N., Gonzalez, C., Jara, M., Alarcón, R., … aurina, E. (2016). Library Resources and Students’ Learning Outcomes: Do All the Resources Have the Same Impact on Learning? The Journal of Academic Librarianship, 42(5), 551–556. https://doi.org/10.1016/J.ACALIB.2016.06.020